In the past, efficient high-flux cooling has been achieved by either wick-type heat pipes or by direct immersion of the device to be cooled in a heat sink liquid. In the latter type of device, circuits are submerged in a fluorocarbon fluid such as freon within a sealed enclosure. As the components dissipate heat, the fluid boils and the vapor bubbles rise to the top of the sealed enclosure to initiate reflux cooling. The vapor is then cooled by extracting heat from the top of the enclosure which causes the vapor to condense and return back to the bottom surrounding the circuits for another heating cycle.
However, each of these types of devices has disadvantages which limits the use thereof in aircraft application in which large switching semiconductors are used, thereby creating a large amount of heat dissipation in a small space.
Wick-type heat pipes are limited by the rate at which cooling fluid can be returned from a condensing zone to a boiling zone adjacent the device being cooled. Conventional wick materials do not achieve sufficiently high heat flux rates to maintain acceptable temperatures in modern electrical power systems with high heat dissipation. Although direct immersion cooling in which the device being cooled causes boiling of a liquid which is later condensed and recycled is very efficient, servicing of the device requires draining the liquid which surrounds it in order to gain access to the device. The device which is cooled by direct immersion must also be packaged in a way which prevents damage from the cooling liquid to the electronic components. This requires extraordinary precautions to make sure that the device is hermetically sealed even after it has been serviced or repaired. Moreover, additional difficulties are encountered in attempting to connect electrical circuits outside the direct immersion sealed enclosure without leakage of the coolant.
Given critical space considerations within aircraft, conventional heat exchangers have also not permitted maximum utilization of space for electrical and electronic packages while, at the same time, permitting adequate cooling. Known heat exchangers are constructed in a way which do not address the problem of configuration. This becomes particularly acute in modern aircraft applications where electronic component power ratings, circuit size and packaging density have increased to address the power requirements for aircraft.
Reflux cooling devices also create internal pressures within the sealed chamber. In the past, this has required greater wall thicknesses and higher overall weight of the reflux enclosure to meet acceptable pressure vessel design criteria. Conventional approaches have not permitted manufacturers to achieve a more compact overall package shape, which is particularly important in aircraft applications. p The use of direct immersion reflux cooling enclosures also has the disadvantage that different circuits immersed in the enclosure will be required to operate at the temperature of the highest heat flux component. This arises from the uniformity of pressure within the vessel which means that all components must operate at the same temperature even if this is not desired for one or more circuits within the enclosure.
Others have attempted to improve upon these two types of cooling arrangements in various ways. In one instance, as shown in Japanese Patent Application 59-217346, a vapor cooling device is provided in which the semiconductor elements to be cooled are disposed externally at the lower part of a liquid reservoir containing a coolant which changes phases between the liquid phase and the vapor phase. As the vaporized coolant rises through a vapor phase pipe 9, it reaches a condensation part where it is condensed between a secondary coolant and then returned to a reservoir through a liquid phase pipe where vapor cooling again begins. However, as a result of this arrangement, it is not possible to stack semiconductor elements in a way which would make the use of the device particularly desirable in the cramped environment of an aircraft. Furthermore, the large internal volume in this arrangement makes it especially undesirable for aircraft use where a small internal volume is necessary because of space considerations in the fuselage, and it does not permit the use of multiple plates to permit different packages from running at different temperatures.
More recently, semiconductor power modules have been proposed in which a heat pipe or capillaries are used to cool the power module. In this connection, U.S. Pat. Nos. 4,727,454 and 4,727,455 are illustrative of these proposals. In one such conventional embodiment, at least one semiconductor power component with a base surface which is soldered to a metallized ceramic substrate. At least one heat conduit or pipe is integrated into the semiconductor power module. The heat conduit includes a condensation area having a larger surface than the base surface over which dissipation heat from the semiconductor power module is distributed. The use of such a module in aircraft applications wherein changes in altitude and "g" forces cause migration of the operating fluid away from the power device to be cooled, thereby decreasing the cooling efficiency, does not provide the overall performance needed in an aircraft electrical power system heat exchanger.
U.S. Pat. No. 4,635,709 attempts to address problems associated with the cooling of electronic devices in aircraft applications. It does so by using air as the primary cooling mode during normal conventional flight. However, a second cooling mode is provided for conditions when air cooling is insufficient or will damage the electronics. This second cooling mode provides a coolant in cold plate grooves. Water or methanol can be used as the coolant which is caused to boil or evaporate as the cold plate heats up. However, this heat exchanger is not intended to work in conjunction with the air flow in order to obtain the cooling benefits of both air cooling and evaporative cooling. Furthermore, it requires that the electronics module be made an integral part of the cooling plate. It also has the disadvantage that the coolant is exhausted overboard and thus is not intended to be recirculated to allow for continuous use, thereby requiring more servicing.